Genotype x Environment Interaction and Yield Stability Analysis of Some Chickpea (Cicer arietinum L.) Genotypes across Different Environments in Sudan

The chickpea or chick pea (Cicer arietinum) is an annuallegume, belongs to familyFabaceae, subfamily Faboideae and genus: Cicer ¹. The plant is a self-pollinating diploid (2n=2x=16) with a 738 Mbp genome ².. At present, the Chickpea has 44 species, of which 35 are perennial, and nine are annual. Chickpea has two market classes—kabuli and desi—based on seed morphology. The plant grows to 20–50 cm (8–20 in) high and has small, feathery leaves on either side of the stem³.

Chickpea (Cicer arietinum L.) crop is the second most important edible legume plant grown worldwide. It is rich, with high nutrition elements,, particularly the carbohydrates (27.42 g.), fat (2.59 g.) which includes saturated fat (0.27 g,), non-saturated fat (0.58 g.) and Polyunsaturated, fat (1.16 g.), Protein (8.86 g. ), Vitamin [VA 0%,Thiamine (V B1. 10%), (VB2. 5 %) , (VB6. 11%.), (VB9. 34 %.), (VB12. 0%.) (VC.2%), (VE2%),(Vk4%]. and Minerals like (Calcium 5%), (Iron 22%,) (Magnesium14%), (Manganese 49%) (Phosphorus 24%,) (Potassium 6%), (Sodium 0%, and (Zinc 16%)⁴. Therefore, Chickpea has an important role in meeting the protein needs of people in under-developing countries, where people need protein for an adequate and balanced nutrition. However protein and vitamin rich foods should have the priority in human diet ⁵. The crop ranks the third among pulses and accounts for 12% of the world pulses production ⁶.

The major contribution (almost 86.73 %) of the world production of chickpea is from Asia, with only 5 % coming from Africa. It is reported that the world area under chickpea cultivation is 13.2 million hectares, with a total production of 11.6 million tons ⁷. The main chickpea producing countries are India, accounting for 67.68 % (7.5 million hectares), and 66.91 % (6.54 million ton) of production followed by Pakistan with 9.75 % (1.08 million hectares and 0.741 million ton. Other chickpea producing countries are Iran, Turkey, Myanmar, Australia, Ethiopia, Canada, Mexico, Syria, USA, Spain, Tanzania and Eritrea ⁸.

In Sudan chickpea is traditionally grown as a winter crop in the northern part, however, its production has expanded recently to the central clay plain of central Sudan. The growing season is restricted to a short period of time by the high temperatures prevailing at the beginning and at the end of the season. Chickpea yield in Sudan varies from 0.83 to 2.8 t/ha, depending on weather conditions genotype and cultural practices ⁹. In Sudan it is an important cash crop which faces strong competition from other winter legume crops, mainly faba bean in its traditional area of production in northern Sudan. In recent years, chickpea area has increased steadily in central Sudan, especially in Gezira scheme and New Half in Eastern Sudan⁹.

Environmental factors, such as soil moisture, sowing time, fertility, temperature and day length have great influences during the various stages of plant growth ¹⁰. Nevertheless, the interaction analysis of G×E is important to identify the superior varieties and their adaptation to and stability in diverse agro-ecological zones ¹¹. However, G×E interaction is defined as a variable phenotypic expression of genotypes in response to environmental changes ¹². This can be performed by the evaluation of genotypes in different environments.

The differences in genotype stability and adaptability to environment can be qualitatively assessed using the biplot graphical representation that scatters the genotypes according to their principal component values ^{13, 14}. To explore the effect of genotype and genotype × environment interaction on grain yield, GGE Bi-plot method is used. Both, AMMI and GGE biplot have frequently been used for explaining GE interaction and to determine high yielding and widely adapted cultivars. These two statistical analyses (AMMI and GGE) have broader relevance for agricultural researchers because they pertain to any two-way data matrices, and such data emerge from many kinds of experiments ¹⁵.

Obviously, low yield is a major problem in the area of Chickpea production. It is caused attributed to several complexes and interacting factors. Therefore, the main objectives of this study are (i) To evaluate the performance of some introduced chickpea genotypes for seed yield and yield related traits under difference environmental conditions in Sudan. (ii) To identify early flowering and fusarium wilt disease resistant genotypes. (iii) To determine the effects of G×E interaction on yield among some chickpea genotypes in order to identify the stable and high yielding genotypes through multiplication yield trials.

The research was conducted the 2017/18 and 2018/19 in four locations under irrigation system. Three studies were conducted in northern region of Sudan. One study was conducted at Shendi Research Station Farm (latitude 16^° 41^' – 16^° 45^' N, longitude 33^° 25^' -33^° 29^'E) in the River Nile State. The second was conducted at Hudeiba Research Station Farm (latitude 17° 33' -17° 37'N, longitude 33^° 56^' - 33^° 59^'E) with an average annual rainfall of 74 mm/ann., while the third study was carried out at Merowe locality, in Research Station Farm, (Latitude: 18^° 27^' - 18^° 31^' N, Longitude: 31° 49' -31° 53' E, Elevation of 258 meters above the sea level). The three Research Station Farms are located in a desert climatic zone, which is characterized by dry and hot climate. In this region, during the winter, temperatures are warm at the day and cool at night, where the temperature can reach to 0°C. The rest of the year, from May to September is very hot, and the temperature could reach to 40 °C, and sometimes could reach up to 50/52°C. Moreover, the wind can raise sandstorms at any time of the year (Sudan Metrology 2022)^{16, 12}. The soil in this region is mostly sand, which is poor due to low nutrient elements and organic matter. except the soil along the river Nile which is mostly loamy clay. The fourth study location was Wad Medani, Gezira State, Gezira Research Station Farm (GRSF) which is located in the central clay plain of Sudan at latitude of 14° 24' N, 14° 36' N, longitude of 33^° 29^' - 33^° 35^' E and elevation of 407 meters above the sea level. Where the soil is heavy cracking clay, alkaline (clay 58%, pH 8.3, organic matter 0.02, nitrogen 0.25, phosphorus 0.06 and potash 3.0%). However, the four locations represent different agro-ecological regions for growing chickpea in Sudan.

Three chickpea genotypes were obtained from the advanced materials of the national chickpea breeding program in the Agricultural Research Corporation (ARC)-Sudan. Moreover, two released chickpea cultivars (Shiekh Mohamed and Burgeig) were used in this study as checks (Table 1).

Across all growing seasons and locations, the land was well prepared using disc plowing, disc- harrowing, leveling and ridging. The experiments were arranged in a Randomized Complete Block Design (RCBD) with four replicates in all locations. Each genotype was planted in a separate plot composed of five ridges; each of 5 m in length, with plant-to-plant and row-to-row distance of 10 and 60 cm, respectively. Sowing was done in the second week of November during the two winter seasons. Frequent irrigation was carried out at 14-16 days intervals to avoid water stress. The experimental plots were treated with a starter dose of nitrogen in the form of urea at the rate of 43kg N/ha before the fourth irrigation. Two hand weeding was done. Seed yield was assessed from a net area of 8.28 m²(3 rows x 4.6 m length x 0.6 m).

During the two seasons, data pertaining the Days to 50% flowering (were registered as days from sowing to the date of flowering, when about 50% of the plants bear at least one flower). Hundred-seed weight (g.) (calculated by 100 seed samples which were randomly selected from each plot], seeds weight (seed weighed using a sensitive balance. The latter parameter was recorded after harvesting, threshing and winnowing (in g or kg)). The seed yield (was weighed using electronic balance on net plot basis and later converted into t/ha.for each genotype).

The five chickpea genotypes were screened for resistance to Fusarium wilt disease in an infested plot at Gezira Research Station Farm during the winter season of 2018/2019, the evaluated for Fusarium wilt disease was done at three stages during the crop cycle (seedling, flowering and pod setting stages). Where disease incidence (%)=(number of wilted plants/ total number of plants) x 100 was determined, and the level of resistance and susceptibility of each tested genotype was determined by using rating scale, with some modifications, where R = 0-20% wilted plants, MR = 21-40% wilted plant, S=41-80% wilted plants and HS = ≥ 80% wilted plants. Furthermore, the wilted plants were checked for vascular discoloration symptoms and re-isolation to confirm the disease was Fusarium wilt caused by fungus Fusarium oxysporum f. sp. Ciceris¹,⁶,¹⁷

The collected data were statistically analyzed using the GenStat 12^th edition¹⁸. The analysis of variance (ANOVA) for measuring all the studied characters was carried out according to the procedure described by ¹⁹.

To determine the performance, stability and genotypic superiority across environments or at specific environment, additive main effect and multiplicative interaction (AMMI) model was used ²⁰ AMMI stability value (ASV) was calculated for each genotype according to the relative contribution of IPCA1 and IPCA2 to the interaction sum square (SS) following the method proposed by ²¹. According to ASV, a genotype with lower ASV score is regarded as more stable. To combine both mean seed yield and stability index in a single criterion, the genotype selection index (GSI) was calculated for each genotype based on the rank of mean seed yield of genotypes across environments and rank of AMMI stability value as proposed by. Similar to ASV, genotype with lower GSI value is regarded as more adapted. Also the GGE biplot model ¹⁹ was followed to test the seed yield stability performance for the three tested chickpea genotypes plus the two checks²².

The results of data analysis for yield and yield related traits in different locations are presented in the Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9 and Table 10 and figure 1 and figure 2.

Considering Table 2, presents the analysis of mean seed yield of the genotypes across the eight environments showing significant effects of genotype, location, year and their interactions on seed yield. The highly significant differences among the tested genotypes indicated that those genotypes which have phenotypic variation and genetic diversity showed effectiveness of selection for the development of new genetic cultivars possessing improved traits. For the average seed yield across the eight environments, the genotype FLIP 08-59 C showed the highest average seed yield and out-yielded the checks, Shiekh Mohamed and Burgeig genotypes in the 6^th- and 5^th- environments, respectively. More-specifically, it out-yielded the check Shiekh Mohamed, in six environments by an average of 20.4% and the second check Burgeig in five environments by an average of 17.4%.

Average seed yield of the five chickpea genotypes across the eight environments varied from 1.45 to/ 1.88 ton/ha. This indicates the wide variability for yield potential among the chickpea genotypes. The three locations, Merowe, Gezira and Shendi, gave the highest seed yield of 2.62, 1.96 and 1.16 ton/ha, respectively. On the other hand, the Hudeiba site recorded the lowest seed yield (0.84 ton/ha). Across all sites, the highest mean seed yield was recorded by the genotype FLIP 08-59,C (1.88 ton/ha) followed by the check Shiekh Mohamed (1.70 ton/ha).

The genotype FLIP 08-59 C produced the higher seed yield compared to the check Burgeig at Medani and Hudeiba sites. The same genotype gave significantly higher seed yield than the check Shiekh Mohamed at Shendi site. At Merowe site, the genotype FLIP 08 -59 C produced a comparable seed yield to the check Burgeig.

At Gezira site, the genotype FLIP 08-59 C produced the highest seed yield and exceeded the two checks, Burgeig and Shiekh Mohamed by 21.1% and 15.8%, respectively. Whereas at Shendi site, the genotype FLIP 08-59 C recorded the highest yield and out-yielded the two checks, Shiekh Mohamed and Burgeig by 35.9% and 11.1%, respectively. At Hudeiba site, also the genotype FLIP 08-59 produced the higher seed yield and exceeded the two checks Burgeig and Shiekh Mohamed by 18.0% and 6.6%, respectively.

The genotype FLIP 09-182 C at Gezira site gave better seed yield and out – yield the two checks Shiekh Mohamed and Burgeig by 8.6% and 14.3%, respectively.

The above Table 3 reflects the combined analysis of variance showing high significant differences (P ≤ 0.001) among the years, location, genotypes, genotype x yea, genotype x location, and genotype x year x location interaction for all studied characters including seed yield Table 4, presents the mean days to 50% flowering of genotypes across eight environments displaying the effect of location and interaction between the genotypes. Year and location were found to be highly significant.

However, across all sites, there were highly significant differences among genotypes in days to 50% flowering with over all mean of days to 50% flowering, ranging between 39 and 48 days in season 2017/18, whereas ranging between 48 and 42 days in 2018/19. The genotype FLIP 08-59 C was the earliest to reach 50% flowering but with non-significant differences as compared to the two checks. Table 5, shows the mean 100-seed weight (g) of genotypes across the eight environments Over all sites, there were highly significant differences among genotypes in 100–seed weight (P ≤ 0.001). The overall mean of 100-seed weight (g) ranged between 19 and 31g. in season 2017/18, whereas, in season 2018/19 the genotype FLIP 09-182 C recorded the heaviest 100-seeds weight (34 g), while the check Burgeig produced the lowest value of 22 g.

Regarding the evaluation of Fusarium wilt disease (Fusarium oxysporumf.sp.ciceris), the results summarized in Table 6 presenting the five chickpea genotypes under disease conditions at Gezira Research Station Farm during the winter of 2018/2019 showed varied resistance to Fusarium wilt at seedling stage and the percent of the infection ranged between 17.2 to 26.5% which is ranked under the resistance level (Resistance to Moderately resistance) of the used scale for the disease. The highest disease incidence was recorded by variety Burgeig (26.5%) followed by genotype FLIP 09-187 C (25.4%) and variety Shiekh Mohamed (23.0%). The lowest disease incidence (17.2%) was scored by the genotype FLIP 08–59 C.

The study showed significant differences among the genotypes on fusarium wilt incidence at flowering and podding stages which ranged from 11.3 to 22.2% and 24.4to 39.8%, From the above results, the genotype FLIP 08–59 C was found to the most resistant to Fusarium wilt disease among all the tested genotypes.

Regarding the Yield Stability, analysis of variance the study showed high significant differences among the tested genotypes (G) for seed yield, environments (E) and their interactions (GEI) Table 7.

The highly significant differences observed among environments indicated that these genotypes were evaluated under diverse seasons and locations. The mean square of G x E interaction also showed high significant differences for most of the traits indicating the effect of the environment and genotype interaction. The mean square of G x E (linear) interaction was not significant.

Stability models were followed to carry out the seed yield stability performance for the three tested chickpea genotypes against the two checks. These were the Additive Main Effect and Multiplicative I nteraction GGE biplot model ²³ and (AMMI) model ²⁴.

Table 8, reflects the Analysis of variance (ANOVA) of AMMI showing the chickpea seed yield was high significantly (P ≤ 0.001) affected by environment (E), genotype (G) and their interaction (GEI). Over all sites, the effect of the environment was significantly, and it was about 90.2% of the total sum of the square, indicating that the environments were diverse. On the other hand, the genotypic effect was explained by significant, but small portion of 3.6% and the (G x E) by (6.3%). The large environmental contribution effect indicated that the maximum variation for the genotypes performance was due to the environmental differences. The significance of the interaction component was 6.3% of the total sum of the square, indicating that the best genotype in one environment is not necessarily the best in another. Therefore, when recommending promising genotypes to an environment we need to take into consider their adaptability and stability.

For the GGE biplot analysis, the GGE refers to the genotype main effect (G) and the genotype x environment interaction (GE) which are the most important sources of variation for a genotype evaluation in multi environmental trials ²⁵. The presence of genotype x environment interaction (GEI) was clearly demonstrated by the AMMI model. When the interaction was partitioned among the first interactions principal component axis (IPCA), they were found to be significant in the assessment. The first principal component (IPCA1) accounted for (58.8%) of the variation caused by interaction, and the second principal component (IPCA2) accounted for (33.2%) of this variation. These are in agreement with the finding of one of the researchers ²⁶. who recommended that the most accurate model for AMMI can be predicted using the first principal components (IPCAs).

Table 9 shows the estimates of stability parameters for yield (t/ha) of five chickpea genotypes in eight environments. AMMI stability values (ASV) revealed variations in yield stability among the 5 chickpea genotypes. This result agrees with the results obtained by one of the studies ²⁷. However stability of a variety is defined as one with ASV value close to zero. Consequently, the G1 (FLIP 08-59C) with ASV value of 0.58 is the most stable after the G4 (Shiekh Mohamed) with ASV value of 0.29. While other genotypes, FLIP 09-182 C, FLIP 09-187 C and check Burgeig are the least stable.

Genotype Selection Index (GSI) measure is essential for quantifying and ranking genotypes according to their yield stability. The one with high seed yield and least (GSI) is considered as the most stable ²⁸. Based on the GSI, the most desirable genotypes for selection of both stability and high seed yielding were the genotypes G4 (Shiekh Mohamed), G1 (FLIP 08-59C) and G2 (FLIP 09-182 C), respectively.

For GGE biplot analysis model, the GGE analysis showed 58.81% and 33.16 % of total variation in the data matrix of GGE respectively, and thus they are accounted for 91.97 % of GGE together (Figure 1).

Figure 1.AMMI 1 biplot of first interaction principal component axis (IPCA1) for mean seed yield t/ha.

According to GGE analysis, the genotypes with PC1 scores is close to zero expressed general adaptation, whereas the larger scores are for more specific adaptation to particular environments ²⁹. Figure 1and Figure 2 showed AMMI and GGE biplots of seed yield of 5 chickpea genotypes across eight environments. The G1 (FLIP 08-59C) was the most stable and best genotypes across the different environments in Sudan.

Figure 2.AMMI 2 biplot of the first interaction principal component axis (IPCA1) versus the second interaction principal component axis (IPCA2) for seed yield t/ha.

The main effects of genotype and environment as well as the G×E were highly significant and the environment caused the most variations in seed yield.

FLIP 08-59 C is found to be an early flowering genotype compared to the two checks. It lso showed good 100 -seed weight and resistant to Fusarium wilt disease, which is considered among the major biotic stresses to the production of chickpea in Sudan.

FLIP 08-59 C produced higher seed yield than all other genotypes overall the environments and performed better at most of the places. It exceeded the two standard checks, Shiekh Mohamed and Burgeig by 9.5% and 10.1%, respectively.

FLIP 08-59 C was found to be comparatively stable as its performance was hardly affected by the G x E interaction and thus would perform well across a wide range of environments. This genotype also was found to be more adapted across the different tested environments. Therefore, the genotype FLIP 08-59 C is recommended for release as a cultivar to be grown at Gezira, River Nile and Northern states of Sudan. Also may be used in future in the breeding program for the development of high yielding and stable genotypes.

1. 1.Science Kew. (2018) . , Plants of the World Online. Retrieved 26.
   Search at Google Scholar

1. 2.R K Varshney, Song C, R K Saxena, Azam S, Yu S et al. (2013) Draft genome sequence of chickpea (Cicer arietinum) provides a resource for trait improvement. , Nat. Biotechnol 31, 240-246.
   View article·Search at Google Scholar

1. 3.E J Knights, K B Hobson. (2016) . Chickpea overview,” in Reference Module in Food Science (2nd Ed 1, 10-1016.
   View article·Search at Google Scholar

1. 4.Gormley Shannon. (2019) (2019)."Little Bean Proves Chickpea Ice Cream Isn't as Weird as It Sounds". Willamette Week. , Retrieved 10.
   Search at Google Scholar

1. 5.Bozoglu H, Ozcelik H. (2005) A Study on determination of genotype x environmental interactions and stabilities of some genotypes in chickpea (CicerarietinumL.).GAP Fourth Agricultural Congress. , Uni. Har. Fac. Agr.,Sanliurfa,1: 834-839.
   Search at Google Scholar

1. 6.Khan R, Farhatullah, Khan H. (2011) Dissection of genetic variability and heritability estimates of chickpea germplasm for various morphological markers and quantitative traits. , Sarhad J. Agric 27(1), 67-72.
   Search at Google Scholar

1. 7.Food. (2012) Agricultural Organization of the United Nations.FASOSTAT. [online] Available athttp://faostat3.org/home/index/html,2012
   Search at Google Scholar

1. 8.Food. (2009) Agricultural Organization of the United Nations.FAOSTAT. [online]. Available athttp://faostat.fao.org/site/567/desktopdefault.aspx?pageid=567,2009
   Search at Google Scholar

1. 9.A T. (1996) Food legume production situation. In: Salih SH, Ageeb OA. Saxena MC, Solh MB (eds) Production and Improvement of Cool-Season Food Legumes in the Sudan. Proceedings of the national research review workshop , Aleppo, Syria: 27-30.
   Search at Google Scholar

1. 10.Tilahun G, Mekbib F, Fikre A, Eshete M. (2015) Genotype × Environment Interaction and Stability Analysis for Yield and Yield Related Traits of Kabuli-Type Chickpea (Cicer arietinumL.) in Ethiopia.Afr. , J. Biotechnol.,14: 1564-1575.
   Search at Google Scholar

1. 11.Kanouni H, Farayedi Y, Saeid A, S H.(2015).Stability analyses for seed yield of chickpea genotypes in the western cold zone of Iran. , Journal of Agricultural Science 7(5), 219-230.
   Search at Google Scholar

1. 12.T, Keneni G, Sefera T, Jarso M, Bekele Y. (2013) Genotype x environment interaction and performance stability for grain yield and Field pea (Pisum sativumL.). , genotypes.Int J. Plant 7(2), 116-123.
   Search at Google Scholar

1. 13.Vita PDe, A M Mastrangeloa, Matteua L, Mazzucotellib E, Virzi N et al. (2010) Genetic improvement effects on yield stability in durum wheat genotypes grown in Italy. Field Crop Res. 119, 68-77.
   Search at Google Scholar

1. 14.Farshadfar E, Zali H, Mohammadi R. (2011) Evaluation of Phenotypic Stability in Chickpea Genotypes Using GGE-Biplot.Ann. , Biol. Res 2, 282-292.
   Search at Google Scholar

1. 15.Naroui M R R, M A Kadir, M Y Rafii, Jaafar Z E Hawa, M R Naghavi et al. (2013) Genotype × environment interaction by AMMI and GGE biplot analysis in three consecutive generations of wheat (Triticum aestivum) under normal and drought stress conditions. , AJS 7(7), 956-961.
   Search at Google Scholar

1. 16.M J Iqbal, Iftikhar K, M B Ilyas. (1993) Evaluation of chickpea germplasm for resistance against wilt disease. , J Agri Res;31: 449-453.
   Search at Google Scholar

1. 17.Suliman W. (2004) Race identification of chickpea wilt pathogen (Fusarium oxysporumf. sp.ciceris) Hudeiba Research Station, Annual Report 2003/04. , Sudan
   Search at Google Scholar

1. 18.GenStat for Windows (2009) . 12thEdition.VSN International Ltd , Hemel Hempstead, UK .
   Search at Google Scholar

1. 19.K A Gomez, A. (1984) Statistical Procedures for Agricultural Research.2nd. edition. , New York 680.
   Search at Google Scholar

1. 20.H G Gauch. (1992) Statistical Analysis of Regional Trials: AMMI Analysis of Factorial Designs. , Amsterdam, The Netherlands 278.
   Search at Google Scholar

1. 21.J L Purchase, H.Hatting and Cs. Vandenventer (2000). Genotype by environments interaction of wheat in South Africa: stability analysis of yield performance. , South Africa J. of Pl. Sci 17, 101-107.
   Search at Google Scholar

1. 22.Farshadfar E. (2008) Incorporation of AMMI stability value and grain yield in a single non-parametric index (GSI) in bread wheat. , Pak J BiolSci 11(14), 1791-1796.
   Search at Google Scholar

1. 23.Yan W. (2002) Singular-value partitioning in biplot analysis of multi-environment trial data.Agron. , J 94(5), 990-996.
   Search at Google Scholar

1. 24.H G Gauch. (1993) Mat model version 2.0. AMMI and related analysis for two-way data matrices.Micro Computer Power. , Ithaca, New York, USA
   Search at Google Scholar

1. 25.Yan W, M S Kang. (2003) GGE biplot analysis: a graphical tool for breeders. , Raton, FL:, In 63-88.
   Search at Google Scholar

1. 26.H G Gauch, R W Zobel. (1996) AMMI analysis of yield trials. In:. , Kang
   Search at Google Scholar

1. 27.J L Purchase. (1997) Parametric analysis to describe G x E interaction and yield stability in winter wheat. Ph.D Thesis, University of the Orange Free State. , Bloemfontein, South Africa
   Search at Google Scholar

1. 28.J S Ebdon, H G Gauch. (2002) Additive Main Effect and Multiplicative Interaction analysis of national turf grass performance trials: I. Interpretation of genotype × environment interaction. , Crop Sci 42, 489-496.
   Search at Google Scholar